You open your analytics dashboard, and the numbers scream: organic traffic is down 30% month-over-month. Your competitor outranks you for a keyword you used to dominate. You suspect technical issues, but a full manual audit takes days — crawling, checking meta tags, analyzing backlinks, mapping redirects. The last time you started one, you gave up by lunch.
Here's the reality: most SEO issues are repetitive and pattern-based. That makes them perfect for AI. But not just any prompt — you need prompts that turn a generic chatbot into a technical SEO auditor, a competitor analyst, and a content strategist. In this guide, I'll share 10 battle-tested prompts that cover the full audit spectrum, from crawling and indexation to semantic core and Core Web Vitals. Each prompt includes a real usage example, so you can copy-paste and adapt it to your own site.
Why AI for SEO Audits?
If you're skeptical, look at the numbers. A 2024 study by BrightEdge found that 68% of online experiences begin with a search engine, and 53% of website traffic comes from organic search. That's a lot of pressure on your site's health. Meanwhile, a typical audit uncovers dozens of issues — broken links, duplicate content, slow pages, missing schema — that are easy to identify but tedious to catalog. AI models like GPT-4, Claude, and Gemini are trained on vast SEO documentation and known best practices, so they can generate comprehensive checklists and even code fixes. The key is to give them the right context and ask for specific, structured output.
Let's dive into the prompts.
1. Crawl Like a Pro: Generate a Technical Audit Checklist
Purpose: Get a personalized, prioritized technical audit checklist based on your site's platform and niche. This is your starting point — a comprehensive map of what to check.
The Prompt:
You are a senior technical SEO auditor with 10+ years of experience. I have a website built on [CMS/platform], targeting the [niche] market. Please generate a prioritized technical SEO audit checklist for my site.
Include the following categories:
- Crawlability and indexation
- On-page elements (title, meta, H1, etc.)
- Site architecture and internal linking
- Page speed and Core Web Vitals
- Mobile-friendliness
- Structured data
- Security and HTTPS
- International SEO (if applicable)
For each item, explain:
- What to check
- Why it matters
- How to fix it if broken
- The potential impact on rankings (e.g., 'high', 'medium', 'low')
Prioritize the items: start with the most critical issues that could cause a significant drop in organic visibility.
Format the output as a Markdown table with columns: Priority, Category, Check, Why It Matters, How to Fix.
Usage Example:
Let's say you have a WordPress e-commerce site selling handmade jewelry. The prompt would return a table that looks like:
| Priority | Category | Check | Why It Matters | How to Fix |
|---|---|---|---|---|
| High | Crawlability | Check robots.txt for accidental disallow of important pages | If /checkout/ is disallowed, search engines can't crawl it, hurting indexation | Remove the Disallow rule, use the robots.txt tester in GSC |
| Medium | On-page | Verify all product pages have unique title tags | Duplicate titles confuse search engines, leading to poor CTR | Use a crawler like Screaming Frog to find duplicates, then rewrite titles |
Why it works: This prompt leverages the AI's knowledge of SEO best practices (like Google's Webmaster Guidelines) and applies them to your specific context. The table format makes it actionable, and the prioritization helps you focus on what matters most.
2. Uncover Crawl Errors and Indexation Issues
Purpose: Diagnose why your pages aren't getting indexed. This prompt helps you analyze common crawl issues like 404s, soft 404s, noindex tags, and canonical problems.
The Prompt:
Act as a technical SEO specialist. I'm providing you with data from my website's log file or crawl (e.g., Screaming Frog export). Here is a sample of URLs and their HTTP status codes:
[Paste your data here]
Analyze this data and identify:
- All 4xx and 5xx errors
- Any redirect chains (more than 3 hops)
- Pages with noindex tags that are still getting traffic
- Duplicate content issues (e.g., similar URLs with slightly different parameters)
For each issue, provide:
- The specific URL or URL pattern
- Why it's a problem for SEO
- A step-by-step fix (e.g., update the .htaccess file, change the canonical tag, etc.)
Also, suggest a list of URLs that should be 301-redirected to preserve link equity.
Usage Example:
You paste a CSV export from Screaming Frog with columns like Address, Status Code, Title, Meta Robots. The AI might find that your /blog/ category has a noindex tag accidentally applied, and it will recommend removing it. Or it might spot a redirect chain from /old-page/ → /new-page/ → /new-page-v2/, costing you link equity.
Why it works: By feeding real data, you ground the AI in your specific situation. The output is a troubleshooting guide you can hand to your dev team.
3. Find and Fix Broken Internal Links
Purpose: Detect broken internal links that hurt user experience and waste crawl budget. This prompt generates a list of links to check and a script to automate the fix.
The Prompt:
You are an SEO analyst. I have a website with the following internal links (from a crawl of my site):
[Paste list of internal links with anchor text and target URLs]
Identify any broken links (links that return 404 or other errors). For each broken link, explain:
- The source page URL
- The anchor text used
- The target URL (if available)
- Why this link is problematic (user experience, crawl budget, link equity loss)
Then, provide a Python script that uses the `requests` library to check all internal links on a given sitemap and output a CSV file with the status codes. Include error handling and rate limiting.
Usage Example:
After running the script, you get a CSV like:
source, target, status, anchor
https://example.com/about, https://example.com/team, 404, Our Team
You can then either update the link or set up a redirect.
Why it works: The prompt combines analysis (identifying broken links) with a practical tool (Python script) that you can reuse. It's a two-in-one solution.
4. Analyze and Improve Core Web Vitals
Purpose: Diagnose Core Web Vitals issues (LCP, INP, CLS) and get actionable optimization tips. This is critical because Core Web Vitals are a ranking factor.
The Prompt:
You are a performance optimization expert. I have a website that is failing Core Web Vitals. Here are my current metrics from PageSpeed Insights:
- LCP: [e.g., 4.2s]
- INP: [e.g., 300ms]
- CLS: [e.g., 0.15]
My website is built on [CMS] and uses [any known technologies, e.g., WordPress, Elementor, and a CDN].
Please provide a detailed analysis of why these metrics might be poor and a step-by-step optimization plan. Cover:
- Server response time (TTFB)
- Image optimization (format, size, lazy loading)
- JavaScript/CSS minification and deferral
- Font loading strategies
- Layout shift causes (e.g., missing dimensions on images, ads)
For each issue, give a concrete recommendation, including code snippets (e.g., a .htaccess rule for caching, a script for lazy loading).
Usage Example:
The AI might recommend switching to WebP images, implementing a caching plugin like WP Rocket, and using font-display: swap. It could even provide a bash script to bulk compress images.
Why it works: It turns vague PageSpeed insights into a concrete plan. The code snippets make it easy to implement.
5. Generate a Semantic Core and Keyword Clustering
Purpose: Build a comprehensive keyword list and organize it into clusters for better content planning. This is a classic SEO task that AI handles beautifully.
The Prompt:
You are a keyword research specialist. I run a website about [topic]. Please generate a list of 100+ long-tail keywords related to my niche, including question-based queries and conversational phrases.
Then, cluster these keywords into logical groups based on search intent (informational, transactional, navigational, commercial). For each cluster, suggest:
- A primary keyword (highest volume and relevance)
- Secondary keywords
- A potential page title and meta description
- A content outline (3-5 main sections)
Use Google's Keyword Planner style: include search volume range (e.g., 1K-10K) and keyword difficulty (low, medium, high). Format as a Markdown table.
Usage Example:
For a site about "vegan recipes," the AI might generate clusters like "vegan dinner recipes," "vegan desserts," "vegan meal prep," each with a list of long-tail keywords like "what can vegans eat for breakfast" or "easy vegan lasagna recipe." It will then give you an outline for a blog post targeting that cluster.
Why it works: It saves hours of manual keyword research and clustering. The AI understands search intent and can mimic the structure of professional SEO tools.
6. Spy on Competitors: Analyze Their Top Pages
Purpose: Understand why your competitors rank well and identify gaps in your content strategy. This prompt helps you reverse-engineer their SEO approach.
The Prompt:
Act as a competitive intelligence analyst. I'm competing with [competitor domain] in the [niche] market. Here is a list of their top-ranking pages (from Ahrefs/SEMrush or a manual search):
[Paste URLs and titles]
Analyze these pages and tell me:
- What keywords they are targeting (from the title, meta, and content)
- The content structure (headings, word count, media)
- Their backlink profile (how many referring domains, any notable guest posts or roundups)
- What makes their content unique (e.g., data, examples, videos)
Then, suggest 5 content ideas for my own site that target similar keywords but offer a different angle or better value. For each idea, provide a working title and a brief outline.
Usage Example:
If your competitor ranks for "best running shoes," the AI might note that their page has a video review and a table of specs. It suggests you create a page with expert input from a podiatrist and a quiz to help users choose. That's a unique angle.
Why it works: It forces you to look at data beyond just rankings. The output is a content strategy you can execute.
7. Recover Lost Backlinks: Outreach Email Generator
Purpose: Create personalized outreach emails to reclaim broken backlinks or get new ones, a key part of off-page SEO.
The Prompt:
You are a digital PR specialist. I've found that my website [your domain] has lost a backlink from [referring domain] (the page is now a 404). I want to reach out to the webmaster to get it fixed.
Here is the context:
- Your site: [your site title and a one-sentence description]
- The referring page URL: [URL]
- The original link context (if known): [e.g., it was a resource page about marketing tools]
Please write a polite, professional email that:
- References the specific link and page
- Explains the 404 issue and how it hurts their users
- Suggests a replacement resource (e.g., your current page or an updated relevant article)
- Includes a call to action
Keep it under 150 words, and use a friendly but authoritative tone.
Usage Example:
The AI generates an email like:
Subject: Quick fix for your resource page
Hi [Name],
I was just checking out your excellent resource list on [topic], and I noticed that the link to our guide on [topic] is now returning a 404. It's a shame because that guide has been updated with [mention new info].
Could you update the link to our new URL: [URL]? It would provide your readers with even more value.
Thanks for your time!
Why it works: Personalized outreach increases response rates. The AI crafts a message that's specific and non-spammy.
8. Automate Redirect Mapping for Site Relaunch
Purpose: When you're moving to a new domain or restructuring URLs, you need a redirect map. This prompt helps you create one and generate the .htaccess code.
The Prompt:
You are an SEO consultant. My website is changing its URL structure. Here is a list of 50 old URLs and their corresponding new URLs:
[Paste mapping]
Please verify that the mapping is logical and complete. Then, generate a set of 301 redirect rules in Apache .htaccess format for these mappings. For any old URL that doesn't have a new equivalent, suggest a sensible redirect (e.g., to the closest category page).
Additionally, provide a checklist of things to do after the redirects are implemented, such as updating internal links, sitemaps, and monitoring for 404s.
Usage Example:
The AI generates a .htaccess block like:
Redirect 301 /old-product-page https://example.com/new-product-page
It might also flag that your /about page has no new equivalent and suggest redirecting it to /about-us.
Why it works: It automates a tedious task and ensures you don't miss any redirects, which is critical for preserving rankings.
9. Optimize Metadata and On-Page Elements at Scale
Purpose: Generate optimized title tags and meta descriptions for all your pages, saving hours of manual writing.
The Prompt:
You are an on-page SEO specialist. I have a list of my website's pages with their current title and meta description. Here is the data:
[Paste table with URL, current title, current meta]
For each page, generate:
- A new title tag (50-60 characters, include primary keyword, and make it compelling)
- A new meta description (150-160 characters, include a call to action)
- A list of 3-5 secondary keywords to include in the content
Make sure the titles are unique and avoid keyword stuffing. Format the output as a CSV-ready table.
Usage Example:
The AI might change your title from "About Us
| Example.com" to "Our Story | Handcrafted Jewelry by Example" and the meta to "Discover the passion behind our handmade jewelry. Shop unique pieces crafted with love and sustainable materials."
Why it works: It scales your optimization efforts. The AI understands character limits and keyword placement.
10. Create a Content Gap Analysis and Action Plan
Purpose: Identify topics you haven't covered but your competitors have, and create a plan to fill those gaps.
The Prompt:
You are a content strategist. I have a list of my top 20 pages and the top 20 pages of my main competitor, [competitor domain]. Here is the data:
[Paste your pages with keywords]
[Paste competitor pages with keywords]
Perform a content gap analysis:
- Identify keywords that your competitor ranks for but you don't
- For each gap, assess the search intent and the difficulty to rank
- Prioritize the gaps based on potential traffic and relevance
Then, create an action plan: for the top 5 gaps, propose a full content brief including target keyword, title, outline, and suggested internal links.
Usage Example:
The AI finds that your competitor ranks for "how to clean silver jewelry," which you haven't covered. It creates a brief with sections like "Common mistakes," "Step-by-step guide," and "Recommended products." You can then assign it to your writer.
Why it works: It's a data-driven approach to content planning. The AI mimics the output of expensive tools like Ahrefs' Content Gap.
Putting It All Together: A Day of AI-Powered Auditing
Imagine starting your morning with Prompt #1 to get a checklist. You spend an hour implementing quick fixes. Then you run Prompt #2 with your crawler data to find indexation issues. By lunch, you've identified your biggest technical problems. Afternoon: you use Prompt #5 to build your semantic core and Prompt #10 to find content gaps. You finish by generating outreach emails (Prompt #7) and new meta tags (Prompt #9). By 5 PM, you have a prioritized list of actions and have already fixed the most critical issues.
Key Takeaway: These prompts aren't magic. They work because they combine your knowledge of your site with the AI's ability to process and structure information. The more context you provide, the better the output.
Pro Tip: Always verify the AI's recommendations against authoritative sources like Google's Search Central documentation. AI can hallucinate, so treat it as a smart assistant, not an oracle.
Now, go open your favorite AI chat interface and start with Prompt #1. Your future self — with a healthier organic traffic — will thank you.
Comments